Lithium battery energy storage system and parameter design method

Abstract

The invention relates to the technical field of lithium batteries, in particular to a lithium battery energy storage system and a parameter design method, and the method comprises the steps: collecting lithium battery operation record fault information and real-time lithium battery loading equipment operation information; constructing a full-parameter knowledge graph recognition network, a related parameter performance recognition network and a single-parameter performance recognition network according to the collected lithium battery operation record fault information; inputting collected real-time operation information of the lithium battery loading equipment into the constructed full-parameter knowledge graph recognition network, the related parameter performance recognition network and the single-parameter performance recognition network, and respectively outputting first performance information, second performance information and third performance information; and predicting the real-time lithium battery according to the first performance information, the second performance information and the third performance information, and calculating a lithium battery energy storage prediction total score of the to-be-predicted equipment. The service life of the lithium battery can be maximized based on the energy storage demand of the lithium battery, and the use efficiency of the lithium battery is improved.

Description

Technical Field

[0001] The present application relates to the technical field of lithium batteries, and particularly relates to a lithium battery energy storage system and a parameter design method thereof. Background Art

[0002] As is well known, the energy storage capacity of a single lithium battery is limited, and it often cannot meet the power supply requirements of electrical energy products when used alone. Since the number of single cells connected in series and parallel is relatively large, the number of electrical connection pieces for series and parallel connection is also relatively large. Improper design of the electrical connection pieces often leads to a reduction in the effective contact area between the cells, resulting in an increase in the resistance value at the connection. When the battery module is powered on, a large amount of heat will be generated at the electrical connection pieces with a large resistance value, resulting in energy loss, which is not conducive to the energy storage heat management and control of the battery module, greatly reducing the energy storage use efficiency of the battery module, shortening the service life of the battery module, and in severe cases, the system temperature of the entire applied battery module will be too high, leading to safety problems.

[0003] At the same time, for equipment loaded with lithium batteries, its energy storage is closely related to its service life. How to maximize the service life of lithium batteries based on the energy storage requirements of lithium batteries is an urgent problem to be solved. Summary of the Invention

[0004] To achieve the above object, the present application provides the following technical solutions:

[0005] According to the first aspect of the present invention, the present invention claims a method for designing lithium battery energy storage parameters, and the method includes the following steps:

[0006] Collect the fault information of the lithium battery operation record and the real-time operation information of the lithium battery loading device;

[0007] According to the collected fault information of the lithium battery operation record, construct a full-parameter knowledge graph recognition network, a related parameter performance recognition network, and a single-parameter performance recognition network;

[0008] Input the collected real-time operation information of the lithium battery loading device into the constructed full-parameter knowledge graph recognition network, related parameter performance recognition network, and single-parameter performance recognition network, and output the first performance information, the second performance information, and the third performance information respectively;

[0009] According to the first performance information, the second performance information, and the third performance information, predict the real-time lithium battery, and calculate the total score of the lithium battery energy storage prediction of the device to be predicted.

[0010] Further, the method further includes the following steps:

[0011] According to the first performance information, the second performance information, and the third performance information, predict the change state of the lithium battery of the device to be predicted, and calculate the change value of the lithium battery of the device to be predicted.

[0012] Further, a method for constructing a full-parameter knowledge graph recognition network based on the collected fault information of the lithium battery operation records includes the following steps:

[0013] Select n full-parameter object information from the lithium battery operation record fault information as a sub-information set, and deploy the sub-information set as the original nodes of a sub-graph;

[0014] Among them, a cycle upper limit is set for the sub-graph;

[0015] Extract a parameter information from the sub-information set according to the requirement, and obtain a boundary point θ according to the requirement;

[0016] The points in the currently selected parameter information that are less than the boundary point θ are arranged on the left pointer of the current node, and the points not less than the boundary point θ are arranged on the right pointer of the current node to form new independent nodes;

[0017] New independent nodes are continuously obtained on the left pointer node and the right pointer node until there is only one parameter information on the independent node, all the features of the parameter information on the node are the same, or the sub-graph has run to the set cycle upper limit.

[0018] Further, the method for constructing a full-parameter knowledge graph recognition network further includes the following steps:

[0019] Calculate the performance scores of all full-parameter object information in the knowledge graph.

[0020] Further, the operation method for the performance scores of all full-parameter object information in the knowledge graph is:

[0021] Calculate the average running trajectory period of a single sub-graph;

[0022] According to the average running trajectory period of a single sub-graph, calculate the performance score of the full-parameter object information x.

[0023] Further, the operation formula for the performance score of the full-parameter object information is:

[0024] s(x,n) = 2 -E(h(x)) / c(n) ;

[0025] Among them, s(x, n) represents the performance score of the full-parameter object information; E(h(x)) represents the expectation of the running trajectory period of the full-parameter object information x in the knowledge graph; h(x) represents the running trajectory period of the full-parameter object information x; the running trajectory period is the number of edges run from the original node of the sub-graph to the independent node.

[0026] Further, a method for constructing a relevant parameter performance recognition network based on the lithium battery operation record fault information includes the following steps:

[0027] Select relevant parameter information tuples from the lithium battery operation record fault information;

[0028] Construct a relevant parameter performance recognition network according to the relevant parameter information tuples.

[0029] Furthermore, constructing a relevant parameter performance recognition network according to the relevant parameter information tuples includes the following sub-steps:

[0030] Calculate the length from the operation loading device object point p to all other loading device object points;

[0031] Calculate the k-th length of the operation loading device object point p

[0032] d k (p) = d(p, o); d(p, o) represents the length between the operation loading device object point O and the operation loading device object point p;

[0033] Collect the k-th length range N k (p) of all operation loading device object points p; the k-th length range includes all operation loading device object points whose length from the operation loading device object point p is not greater than the k-th length;

[0034] Calculate the effective length of the operation loading device object point p;

[0035] Calculate the specified effective life of all operation loading device object points according to the effective length of the operation loading device object point p;

[0036] Calculate the specified weight coefficient of the operation loading device object point according to the specified effective life.

[0037] Furthermore, constructing a single-parameter change upper limit network according to the lithium battery operation record fault information includes the following steps:

[0038] Select single-parameter object information from the lithium battery operation record fault information;

[0039] Define the parameter information time series of the single-parameter object information as y i , and define the number of time instants as x i , and construct a deep learning network according to the parameter information time series;

[0040] Calculate the change upper limit range of the parameter information at the next moment according to the deep learning network.

[0041] According to the second aspect of the present invention, the present invention claims protection for a lithium battery energy storage system, which includes:

[0042] An information acquisition system for acquiring lithium battery operation record fault information and real-time lithium battery loading device operation information;

[0043] A network construction unit for constructing a full-parameter knowledge graph recognition network, a related-parameter performance recognition network, and a single-parameter performance recognition network based on the collected fault information of the lithium battery operation records.

[0044] A performance information acquisition unit for inputting the collected real-time lithium battery loading device operation information into the constructed full-parameter knowledge graph recognition network, related-parameter performance recognition network, and single-parameter performance recognition network, and respectively acquiring the first performance information, the second performance information, and the third performance information.

[0045] An information processor for predicting the real-time lithium battery according to the first performance information, the second performance information, and the third performance information, and calculating the total lithium battery energy storage prediction score of the device to be predicted.

[0046] This application relates to the technical field of lithium batteries, and in particular to a lithium battery energy storage system and a parameter design method. It collects the fault information of the lithium battery operation records and the real-time lithium battery loading device operation information; constructs a full-parameter knowledge graph recognition network, a related-parameter performance recognition network, and a single-parameter performance recognition network according to the collected fault information of the lithium battery operation records; inputs the collected real-time lithium battery loading device operation information into the constructed full-parameter knowledge graph recognition network, related-parameter performance recognition network, and single-parameter performance recognition network, and respectively outputs the first performance information, the second performance information, and the third performance information; predicts the real-time lithium battery according to the first performance information, the second performance information, and the third performance information, and calculates the total lithium battery energy storage prediction score of the device to be predicted. The present invention can maximize the service life of the lithium battery based on the energy storage demand of the lithium battery and improve the use efficiency of the lithium battery. Description of the Drawings

[0047] Figure 1 It is a working flowchart of a lithium battery energy storage parameter design method requested to be protected by an embodiment of the present application.

[0048] Figure 2 It is a second working flowchart of a lithium battery energy storage parameter design method requested to be protected by an embodiment of the present application.

[0049] Figure 3 It is a third working flowchart of a lithium battery energy storage parameter design method requested to be protected by an embodiment of the present application.

[0050] Figure 4 It is a structural module diagram of a lithium battery energy storage system requested to be protected by an embodiment of the present application. Detailed Embodiments

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0052] The terms "first", "second", and "third" in the present application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the number of the indicated technical features. Thus, the features defined with "first", "second", and "third" may explicitly or implicitly include at least one of such features. In the description of the present application, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present application are only used to explain the relative positional relationship and motion state between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.

[0053] Referring to "embodiments" in this article means that the specific features, structures, or characteristics described in connection with the embodiments may be included in at least one embodiment of the present application. The phrase appears in various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0054] Embodiment 1

[0055] As Figure 1 shown, the present application provides a method for designing lithium battery energy storage parameters, and the method includes the following steps:

[0056] Step S1, collect the lithium battery operation record fault information and the real-time lithium battery loading device operation information.

[0057] Among them, the lithium battery operation record fault information includes the information of recording the lithium battery loading device operation information and the information of recording the lithium battery parameter acquisition information. The information of recording the lithium battery loading device operation information is the information of automatically operating the lithium battery by using an automatic lithium battery operation device; the information of recording the lithium battery parameter acquisition information is the information obtained by collecting the lithium battery in a certain way.

[0058] Step S2: Construct a full-parameter knowledge graph recognition network, a related-parameter performance recognition network, and a single-parameter performance recognition network based on the collected fault information of the lithium battery operation record.

[0059] Specifically, construct a full-parameter knowledge graph network based on the full-parameter information of the lithium battery operation record fault information. The full-parameter information includes all information of multiple parameters. For example, the full-parameter information is a tuple of parameter information.

[0060] As Figure 2 shown, as a specific embodiment of the present invention, the method for constructing a full-parameter knowledge graph recognition network based on the collected fault information of the lithium battery operation record includes the following steps:

[0061] Step S210: Select n full-parameter object information from the lithium battery operation record fault information as a sub-information set, and deploy the sub-information set as the original nodes of a sub-graph.

[0062] Preferably, the upper limit of the period of the sub-graph is

[0063] h = ceiling(log2n); where h represents the period;

[0064] ceiling() represents the maximum period function; n represents the total number of selected full-parameter object information.

[0065] Among them, the sub-information set includes multiple full-parameter information, and each full-parameter information includes multiple parameter information.

[0066] Step S220: Extract a parameter information according to the requirement from the sub-information set, and obtain a boundary point θ according to the requirement.

[0067] Specifically, within the information range of the current sub-graph nodes, obtain a boundary point θ according to the requirement. The boundary point is a number between the maximum value and the minimum value of the parameter information in the current sub-information set.

[0068] Step S230: Arrange the points less than the boundary point among the currently selected parameter information on the left pointer of the current node, and arrange the points not less than the boundary point on the right pointer of the current node to form new independent nodes.

[0069] Step S240: Recursively perform steps S220 and S230 on the left pointer node and the right pointer node to continuously obtain new independent nodes until there is only one parameter information on the independent node, all the features of the parameter information on the node are the same, or the sub-graph has run to the set period upper limit h.

[0070] Step S250: Calculate the performance scores of all full-parameter object information in the knowledge graph.

[0071] Step S250 includes the following sub-steps:

[0072] Step S251, calculate the average running trajectory period of a single sub-graph.

[0073] Specifically, the calculation formula for the average running trajectory period of a single sub-graph is:

[0074]

[0075] Where c(n) represents the average running trajectory period of the sub-graph; H(n - 1) represents the harmonic number, and n represents the total number of full-parameter object information selected.

[0076] Step S252, calculate the performance score of the full-parameter object information according to the average running trajectory period of a single sub-graph.

[0077] Specifically, the calculation formula for the performance score of the full-parameter object information x is:

[0078] s(x,n) = 2 -E(h(x)) / c(n) ;

[0079] Where s(x, n) represents the performance score of the full-parameter object information x; E(h(x)) represents the expectation of the running trajectory period of the full-parameter object information x in the knowledge graph; h(x) represents the running trajectory period of the full-parameter object information x; the running trajectory period is the number of edges traveled from the original node to the independent node of the sub-graph.

[0080] As a specific embodiment of the present invention, the method for constructing a relevant parameter performance recognition network based on the lithium battery operation record fault information includes the following steps:

[0081] Step S310, select relevant parameter information tuples from the lithium battery operation record fault information.

[0082] As Figure 3 shown, step S310 includes the following sub-steps:

[0083] Step S311, select the lithium battery parameter acquisition information of the lithium battery operation station in the most recent year from the lithium battery operation record fault information.

[0084] Step S312, initialize each lithium battery parameter acquisition information to form a new parameter sequence.

[0085] Step S313, calculate the grey correlation degree between all lithium battery parameter acquisition information.

[0086] Step S314, extract the lithium battery parameter acquisition information with a correlation degree greater than 0.7 as the lithium battery relevant parameter information.

[0087] Step S315: According to the relevant parameter information of the lithium battery, collect the operating information of the lithium battery loading device corresponding to each relevant parameter of the lithium battery, and form an information set as the relevant parameter information tuple.

[0088] Step S320: Construct a relevant parameter performance recognition network according to the relevant parameter information tuple.

[0089] Step S320 includes the following sub-steps:

[0090] Step S321: Calculate the length from the loading device object point p to all other loading device object points. For example: extract the calculation method.

[0091] Among them, the loading device object point refers to the object information of the operating information of the lithium battery loading device.

[0092] Step S322: Calculate the l-th length d

[0093] d k (p) = d(p, o) and satisfies the following conditions:

[0094] There are at least k points in the tuple that do not include the loading device object point p.

[0095] o′ ∈ C(x ≠ p), satisfying

[0096] d(p, o′) ≤ d(p, o);

[0097] There are at most k - 1 points in the tuple that do not include the loading device object point p.

[0098] o′ ∈ C(x ≠ p), satisfying

[0099] d(p, o′) < d(p, o).

[0100] Step S323: Collect the k-th length range N k (p) of all loading device object points p, and this k-th length range includes all loading device object points whose length with the loading device object point p is not greater than the k-th length.

[0101] Step S324: Calculate the effective length of the loading device object point p. The effective length between the loading device object point o and the loading device object point p is defined as:

[0102] reach-d k (p, o) = max{d k (o), d(p, o)};

[0103] Among them, reach-d k(p, o) represents the effective length between the loading device object point o and the loading device object point p; max{} represents taking the maximum value; d k (o) represents the k-th length of the loading device object point o; d(p, o) represents the actual length between the loading device object point o and the loading device object point p.

[0104] Step S325: Calculate the specified effective life of all loading device object points according to the effective length of the loading device object points.

[0105] The calculation formula for the specified effective life of the loading device object points is:

[0106]

[0107] Among them, lrd k (p) represents the specified effective life of the loading device object point; N k (p) represents the k-th length range of the loading device object point p; reach-d k (p, o) represents the effective length between the loading device object point o and the loading device object point p.

[0108] Step S326: Calculate the specified weight coefficient of the loading device object point according to the specified effective life.

[0109] The calculation formula for the specified weight coefficient of the loading device object points is:

[0110]

[0111] Among them, LOF k (p) represents the specified weight coefficient of the loading device object point; N k (p) represents the k-th length range of the loading device object point p; lrd k (o) represents the specified effective life of the loading device object point O; lrd k (p) represents the specified effective life of the loading device object point.

[0112] If the specified weight coefficient is closer to 1, it indicates that the life of the loading device object point p is similar to that of its range point, and the loading device object point p may belong to the same cluster as its range; if the specified weight coefficient is less than 1, it indicates that the life of the loading device object point p is higher than that of its range point, and the loading device object point p is a dense point; if the specified weight coefficient is greater than 1, it indicates that the life of the loading device object point p is less than that of its range point, and the loading device object point p is more likely to be a performance point.

[0113] As a specific embodiment of the present invention, according to the fault information of the lithium battery operation record, a single-parameter change upper limit network is constructed, including the following steps:

[0114] Step T1, select single-parameter object information from the fault information recorded during the operation of the lithium battery.

[0115] Specifically, eliminate the performance information from the fault information recorded during the operation of the lithium battery, and select the single-parameter object information.

[0116] Step T1 includes the following steps:

[0117] Step T110, for each parameter information, calculate the average value of the first 12 pieces of lithium battery parameter acquisition information of the current section as the first average value information.

[0118] Step T120, based on the operation information of the lithium battery loading equipment, calculate the average value of the operation information of the lithium battery loading equipment in the previous three months as the first life average value information, and the average value of the operation information of the lithium battery loading equipment in the previous month as the second life average value information.

[0119] Specifically, collect the operation information of the lithium battery loading equipment in the previous three months after eliminating the performance information, and calculate the average value of the collected operation information of the lithium battery loading equipment in the previous three months as the first life average value information; collect the operation information of the lithium battery loading equipment in the previous month after eliminating the performance information, and calculate the average value of the collected operation information of the lithium battery loading equipment in the previous month as the second life average value information.

[0120] Step T130, combine the obtained first average value information, first life average value information, and second life average value information with the parameter information at multiple moments before the real time to form the parameter information time series of each parameter as the single-parameter object information.

[0121] The parameter information time series includes the parameter information at multiple moments.

[0122] Step T2, define the parameter information time series of the single-parameter object information as y i , define the number of moments as x i , and construct a deep learning network according to the parameter information time series.

[0123] The deep learning network is:

[0124] Among them, i is the parameter identifier, x i is the number of moments;

[0125] Input the parameter information time series y i and the number of moments x i into the deep learning network to obtain the values of a and b.

[0126] Step T3, according to the deep learning network, calculate the upper limit range a of the change in the parameter information at the next moment i .

[0127] Specifically, the operation formula for the upper limit range of variation is as follows:

[0128]

[0129] Wherein, is the operation value of the deep learning network at the next moment, and y i is the parameter information at the next moment; std() represents the standard deviation.

[0130] Step S3: Input the collected real-time operation information of the lithium battery loading device into the constructed full-parameter knowledge graph recognition network, relevant parameter performance recognition network, and single-parameter performance recognition network, and output the first performance information, second performance information, and third performance information respectively.

[0131] As a specific embodiment of the present invention, the method for collecting the first performance information is as follows:

[0132] Input the real-time operation information of the lithium battery loading device into the full-parameter knowledge graph recognition network, calculate the lengths of the real-time operation information of the lithium battery loading device and all recorded full-parameter object information, collect the performance scores of the object information with the closest length, and use the performance scores of the object information with the closest length as the performance scores of the real-time operation information of the lithium battery loading device. If the performance scores of the real-time operation information of the lithium battery loading device are close to 1, then regard the real-time operation information of the lithium battery loading device as the first performance information; otherwise, regard it as invalid information.

[0133] As a specific embodiment of the present invention, the method for collecting the second performance information is as follows:

[0134] Input the real-time operation information of the lithium battery loading device into the relevant parameter performance recognition network, calculate the specified weight coefficient of the loading device object point. If the specified weight coefficient of the loading device object point exceeds the preset upper limit, then define the information of the loading device object point as the second performance information; otherwise, it is invalid information.

[0135] As a specific embodiment of the present invention, the method for collecting the third performance information is as follows:

[0136] Input the real-time operation information of the lithium battery loading device into the single-parameter performance recognition network. If the parameter information y i of the real-time operation information of the lithium battery loading device is not within the variation upper limit range a i , then define the parameter information of the real-time operation information of the lithium battery loading device as the third performance information; otherwise, it is invalid information.

[0137] Step S4: According to the first performance information, second performance information, and third performance information, predict the real-time lithium battery, and calculate the total score of the lithium battery energy storage prediction of the device to be predicted.

[0138] Step S4 includes the following sub-steps:

[0139] Step S410: Collect the first performance information, second performance information, and third performance information of all lithium battery operation sites within the range of the device to be predicted during a preset time period.

[0140] Step S420: Calculate the total score of the lithium battery energy storage prediction of the device to be predicted based on the collected first performance information, second performance information, and third performance information.

[0141] Specifically, the calculation formula for the total score of the lithium battery energy storage prediction of the device to be predicted is as follows:

[0142]

[0143] Among them, Y represents the total score of the lithium battery energy storage prediction of the device to be predicted; α1 represents the accuracy coefficient for collecting the first performance information; α2 represents the accuracy coefficient for collecting the second performance information; α3 represents the accuracy coefficient for collecting the third performance information; M represents the total number of the first performance information collected; Q represents the total number of the second performance information collected; D represents the total number of the third performance information collected; Z represents the total number of all lithium battery operation sites; i represents the collection of the i-th first performance information; j represents the j-th parameter information in the first performance information; N represents the total number of types of parameter information in the first performance information; δ ij represents the energy storage coefficient of the j-th parameter information in the i-th first performance information; U1 ij represents the measured value of the j-th parameter information in the i-th first performance information; V1 ij represents the standard value of the j-th parameter information in the i-th first performance information; q represents the q-th second performance information; e represents the e-th parameter information in the second performance information; E represents the total number of types of parameter information in the second performance information; η qe represents the energy storage coefficient of the e-th parameter information in the q-th second performance information; U2 qe represents the measured value of the e-th parameter information in the q-th second performance information; V2 qe represents the standard value of the e-th parameter information in the q-th second performance information; φ d represents the energy storage coefficient of the d-th third performance information; U3 d represents the measured value of the d-th third performance information; V3 d represents the standard value of the d-th third performance information; G 12 represents the total number of parameter information that is repeated among the first performance information, second performance information, and third performance information.

[0144] Among them, the first performance information is full parameter information, the second performance information is relevant parameter information, and the third performance information is single parameter information.

[0145] Step S5: Predict the change state of the lithium battery of the device to be predicted according to the first performance information, the second performance information, and the third performance information, and calculate the change value of the lithium battery of the device to be predicted.

[0146] Specifically, the calculation formula for the change value of the lithium battery of the device to be predicted is:

[0147]

[0148] Among them, SG represents the change value of the lithium battery of the device to be predicted; t represents the t-th moment; T represents the total number of moments for collecting lithium battery information; α1 represents the accuracy coefficient for collecting the first performance information; α2 represents the accuracy coefficient for collecting the second performance information; α3 represents the accuracy coefficient for collecting the third performance information; M represents the total number of the first performance information collected; N represents the total number of types of parameter information in the first performance information; Q represents the total number of the second performance information collected; E represents the total number of types of parameter information in the second performance information; D represents the total number of the third performance information collected; represents the weight of the j-th parameter information in the i-th first performance information; represents the weight of the e-th parameter information in the q-th second performance information; represents the weight of the parameter information in the d-th third performance information; represents the measured value of the j-th parameter information in the i-th first performance information collected at the (t + 1)-th moment; represents the measured value of the j-th parameter information in the i-th first performance information collected at the t-th moment; represents the measured value of the e-th parameter information in the q-th second performance information collected at the (t + 1)-th moment; represents the measured value of the e-th parameter information in the q-th second performance information collected at the t-th moment; represents the measured value of the d-th third performance information collected at the (t + 1)-th moment; represents the measured value of the d-th third performance information collected at the t-th moment.

[0149] Embodiment 2

[0150] As Figure 4 shown, the present application also provides a lithium battery energy storage system 100, and this system includes:

[0151] An information collection system 10, which is used to collect lithium battery operation record fault information and real-time lithium battery loading equipment operation information;

[0152] The network construction unit 20 is used to construct a full-parameter knowledge graph recognition network, a related-parameter performance recognition network, and a single-parameter performance recognition network according to the collected fault information of the lithium battery operation records.

[0153] The performance information collection unit 30 is used to input the collected real-time lithium battery loading device operation information into the constructed full-parameter knowledge graph recognition network, related-parameter performance recognition network, and single-parameter performance recognition network, and output the first performance information, the second performance information, and the third performance information respectively.

[0154] The information processor 40 is used to predict the real-time lithium battery according to the first performance information, the second performance information, and the third performance information, and calculate the total score of the lithium battery energy storage prediction of the device to be predicted.

[0155] The information processor 40 is also used to predict the change state of the lithium battery of the device to be predicted according to the first performance information, the second performance information, and the third performance information, and calculate the change value of the lithium battery of the device to be predicted.

[0156] Among them, the operation formula of the total score of the lithium battery energy storage prediction of the device to be predicted is as follows:

[0157]

[0158] Among them, Y represents the total score of the lithium battery energy storage prediction of the device to be predicted; α1 represents the accuracy coefficient of collecting the first performance information; α2 represents the accuracy coefficient of collecting the second performance information; α3 represents the accuracy coefficient of collecting the third performance information; M represents the total number of the first performance information collected; Q represents the total number of the second performance information collected; D represents the total number of the third performance information collected; Z represents the total number of all lithium battery operation sites; i represents the collection of the i-th first performance information; j represents the j-th parameter information in the first performance information; N represents the total number of types of parameter information in the first performance information; δ ij represents the energy storage coefficient of the j-th parameter information in the i-th first performance information; U1 ij represents the measured value of the j-th parameter information in the i-th first performance information; V1 ij represents the standard value of the j-th parameter information in the i-th first performance information; q represents the q-th second performance information; e represents the e-th parameter information in the second performance information; E represents the total number of types of parameter information in the second performance information; η qe represents the energy storage coefficient of the e-th parameter information in the q-th second performance information; U2 qe represents the measured value of the e-th parameter information in the q-th second performance information; V2 qe represents the standard value of the e-th parameter information in the q-th second performance information; φ dThe energy storage coefficient representing the d-th third performance information; U3 d The measured value representing the d-th third performance information; V3 d The standard value representing the d-th third performance information; G 12 Represents the total number of parameter information that is repeated among the first performance information, the second performance information, and the third performance information.

[0159] Among them, the calculation formula for the change value of the lithium battery of the device to be predicted is:

[0160]

[0161] Among them, SG represents the change value of the lithium battery of the device to be predicted; t represents the t-th moment; T represents the total number of moments for collecting lithium battery information; α1 represents the accuracy coefficient for collecting the first performance information; α2 represents the accuracy coefficient for collecting the second performance information; α3 represents the accuracy coefficient for collecting the third performance information; M represents the total number of the first performance information collected; N represents the total number of types of parameter information in the first performance information; Q represents the total number of the second performance information collected; E represents the total number of types of parameter information in the second performance information; D represents the total number of the third performance information collected; Represents the weight of the j-th parameter information in the i-th first performance information; Represents the weight of the e-th parameter information in the q-th second performance information; Represents the weight of the parameter information in the d-th third performance information; Represents the measured value of the j-th parameter information in the i-th first performance information collected at the (t + 1)-th moment; Represents the measured value of the j-th parameter information in the i-th first performance information collected at the t-th moment; Represents the measured value of the e-th parameter information in the q-th second performance information collected at the (t + 1)-th moment; Represents the measured value of the e-th parameter information in the q-th second performance information collected at the t-th moment; Represents the measured value of the d-th third performance information collected at the (t + 1)-th moment; Represents the measured value of the d-th third performance information collected at the t-th moment.

[0162] Among them, the first performance information is full parameter information, the second performance information is relevant parameter information, and the third performance information is single parameter information.

[0163] The beneficial effects achieved by this application are as follows:

[0164] (1) This application constructs a multi-level performance value recognition method for multi-parameters, related parameters, and single parameters, effectively improving the recognition accuracy of performance values in operation information, reducing the false detection rate and missed detection rate of performance value recognition, ensuring the authenticity and objectivity of operation information, and providing information support for early detection of impending low energy operation of lithium batteries.

[0165] (2) This application predicts the lithium battery of the device to be predicted and predicts the change state of the lithium battery based on the collected first performance information, second performance information, and third performance information, so as to achieve a comprehensive prediction of the lithium battery state. According to the functional use of the device to be predicted, it is judged whether the current lithium battery state of the lithium battery to be predicted meets its functional use, and according to the change state of the lithium battery, it is judged whether the change of the lithium battery meets the requirements, so as to achieve better management of the lithium battery state.

[0166] In several embodiments provided by this application, it should be understood that the disclosed systems, systems, and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the system or unit can be in electrical, mechanical or other forms.

[0167] In addition, each functional unit in various embodiments of this application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. The above is only the implementation mode of this application, and it does not limit the patent scope of this application. Any equivalent structure or equivalent process transformation made by using the specification and drawings of this application, or directly or indirectly applied in other related technical fields, is equally included in the patent protection scope of this application.

[0168] The specific implementation mode of the invention has been described in detail above, but it is only an example, and this application is not limited to the specific implementation mode described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of this application. Therefore, equal transformation, modification, improvement, etc. made without departing from the spirit and principle of this application should all be covered within the scope of this application.

Claims

1. A lithium battery energy storage parameter design method, characterized in that: The method comprises the following steps: Collect lithium battery operation record fault information and real-time lithium battery loading equipment operation information; According to the collected lithium battery operation record fault information, a full-parameter knowledge graph recognition network, a related parameter performance recognition network and a single-parameter performance recognition network are constructed; The collected real-time lithium battery loading equipment operation information is input into the constructed full-parameter knowledge graph recognition network, related parameter performance recognition network and single-parameter performance recognition network, and the first performance information, the second performance information and the third performance information are output respectively; The real-time lithium battery is predicted based on the first performance information, the second performance information and the third performance information, and the total score of the lithium battery energy storage prediction of the device to be predicted is calculated.

2. The lithium battery energy storage parameter design method according to claim 1, characterized in that: The method further comprises the following steps: The change state of the lithium battery of the device to be predicted is predicted according to the first performance information, the second performance information and the third performance information, and the change value of the lithium battery of the device to be predicted is calculated.

3. The lithium battery energy storage parameter design method according to claim 1, characterized in that: According to the collected lithium battery operation record fault information, the method for constructing a full-parameter knowledge graph recognition network includes the following steps: From the lithium battery operation record fault information, select n full parameter object information as a sub-information set, and deploy the sub-information set to an original node of a sub-graph; Among them, an upper limit of the period is set for the sub-graph; Extract a parameter information from the sub-information set according to the requirements, and obtain a boundary point θ according to the requirements; Place the points in the currently selected parameter information that are smaller than the boundary point θ at the left pointer of the current node, and the points that are not smaller than the boundary point θ at the right pointer of the current node to form a new independent node; New independent nodes are continuously obtained at the left pointer node and the right pointer node until there is only one parameter information on the independent node, all features of the parameter information on the node are consistent, or the sub-graph has run to the set cycle upper limit.

4. The lithium battery energy storage parameter design method according to claim 3, characterized in that: The method for constructing a full-parameter knowledge graph recognition network also includes the following steps: The performance score of computing all full-parameter object information in the knowledge graph.

5. The lithium battery energy storage parameter design method according to claim 4, characterized in that: The calculation method for the performance score of all full-parameter object information in the knowledge graph is: Calculate the average trajectory period of a single sub-graph; The performance score of the full parameter object information x is calculated based on the average trajectory period of a single sub-graph.

6. The lithium battery energy storage parameter design method according to claim 5, characterized in that: The performance score calculation formula for all parameter object information is: S(x,n)=2 -E(h(x)) / c(n) ; Among them, s(x, n) represents the performance score of the full-parameter object information; E(h(x)) represents the expectation of the running trajectory period of the full-parameter object information x in the knowledge graph; h(x) represents the running trajectory period of the full-parameter object information x; the running trajectory period is the number of edges running from the original node of the subgraph to the independent node.

7. The lithium battery energy storage parameter design method according to claim 1, characterized in that: According to the fault information recorded in the operation of the lithium battery, the method for constructing a relevant parameter performance identification network includes the following steps: Select relevant parameter information tuples from lithium battery operation record fault information; According to the relevant parameter information tuple, a relevant parameter performance identification network is constructed.

8. The lithium battery energy storage parameter design method according to claim 7, characterized in that: According to the relevant parameter information tuple, constructing the relevant parameter performance identification network includes the following sub-steps: Calculate the length from the loading device object point p to all other loading device object points; Calculate the kth length of the loading device object point p d k (p) = d(p, o); d(p, o) represents the length between the loading device object point o and the loading device object point p; Collect the k-th length range N of all loading equipment object points p k (p); the k-th length range includes all loading device object points whose length with the loading device object point p is not greater than the k-th length; Calculate the effective length of the loading device object point p; According to the effective length of the loading device object point p, the specified effective life of all loading device object points is calculated; Calculates the specified weight coefficient of the loading equipment object point based on the specified effective life.

9. The lithium battery energy storage parameter design method according to claim 1, characterized in that: According to the fault information recorded in the operation of lithium batteries, a single parameter change upper limit network is constructed, including the following steps: Select single parameter object information from lithium battery operation record fault information; Define the parameter information sequence of the single parameter object information as y i , the time number is defined as x i , build a deep learning network based on the time series of parameter information; According to the deep learning network, the upper limit range of the change of parameter information at the next moment is calculated.

10. A lithium battery energy storage system, characterized in that: The system includes: Information collection system, used to collect lithium battery operation record fault information and real-time lithium battery loading equipment operation information; A network construction unit, used to construct a full-parameter knowledge graph recognition network, a related parameter performance recognition network and a single-parameter performance recognition network according to the collected lithium battery operation record fault information; A performance information collection unit is used to input the collected real-time lithium battery loading equipment operation information into the constructed full-parameter knowledge graph recognition network, the related parameter performance recognition network and the single-parameter performance recognition network, and respectively collect the first performance information, the second performance information and the third performance information; The information processor is used to predict the real-time lithium battery according to the first performance information, the second performance information and the third performance information, and calculate the total score of the lithium battery energy storage prediction of the device to be predicted.

Images